How to load state_dict for ParameterList?

1

I try to load pretrained model by:

saved = torch.save(model.state_dict())
model.load_state_dict(torch.load(saved))

but the ParameterList([Parameter(Tensor1), Parameter(Tensor2) …]) of the model is not correctly loaded.

So, what is the correct method to load pretrained weights for nn.ParameterList()?

pytorch 1.6.0

2

This small code snippet seems to work, so could you share a minimal code example to reproduce this issue?

# create a simple model using a linear layer
class MyModel(nn.Module):
    def __init__(self):
        super(MyModel, self).__init__()
        self.params = nn.ParameterList([
            nn.Parameter(torch.randn(1)),
            nn.Parameter(torch.randn(1)),
            nn.Parameter(torch.randn(1))
        ])
        self.fc1 = nn.Linear(1, 1)
        
    def forward(self, x):
        x =self.fc1(x)
        return x

model = MyModel()
print(model.state_dict())

sd = model.state_dict()
model = MyModel()
model.load_state_dict(sd)
print(model.state_dict())